Research Area:  Machine Learning
Achieving accurate stock market models can provide investors with tools for making better data-based decisions. These models can help traders to reduce investment risk and select the most profitable stocks. Furthermore, creating advanced models enable the usage of non-traditional data like historical stock prices and news. There are several review articles about financial problems, including stock market analysis and forecast, currency exchange forecast, optimal portfolio selection, among others. However, the recent advances in machine learning techniques, like Deep Learning, Text Mining Techniques, and Ensemble Techniques, raises the need to perform an updated review. This study aims to fill this gap by providing an updated systematic review of the forecasting techniques used in the stock market, including their classification, characterization and comparison. The review is focused on studies on stock market movement prediction from 2014 to 2018, obtained from the scientific databases Scopus and Web of Science. Besides, it analyzes surveys and other reviews of recent studies published in the same time frame and the same databases.
Author(s) Name:  O Bustos, A. Pomares-Quimbaya
Journal name:  Expert Systems with Applications
Publisher name:  Elsevier
Volume Information:  Volume 156, 15 October 2020, 113464
Paper Link:   https://www.sciencedirect.com/science/article/abs/pii/S0957417420302888